Demand Analysis

1

OPERATIONS PLANNING

The operations planning considers how to optimize BRT ridership, capacity, and service frequency. It involves understanding the existing travel demand and forecasting future demand of the proposed BRT services. The network of BRT routes is optimized and details of the BRT service plan are resolved, including BRT stop spacing, speed and capacity, and impacts on general traffic.6

DEMAND ANALYSIS

Travel demand analysis involves considering

where the existing public transport demand is concentrated and then

estimating where there is potential future BRT demand.

Some of the existing bus and rail demand between Everett, Boston, and surrounding areas is expected to transfer to a new BRT corridor. Some trips currently made on other modes may be drawn to the BRT, and finally the BRT services may induce some new BRT trips from Everett to new destinations. These elements combine to make up the forecasted BRT demand.

The existing public transit demand is the most important factor in determining whether a corridor is appropriate for BRT.7 The existing transit riders using the corridor are most likely going to benefit from the BRT. Analysis of the potential BRT passenger demand between Everett and Boston is the foundation for subsequent corridor planning and design. The demand analysis will inform the infrastructure design, ensuring adequate capacity to efficiently handle future growth in ridership. The demand estimate will also dictate the vehicle size, fleet size, and BRT service plan.

DEVELOP AND REFINE EVERETT-BOSTON BRT DEMAND ESTIMATES

There are six main steps to developing and refining the demand analysis for BRT between Everett and Boston:

1

Understand the existing public transport network;

Create a rapid demand assessment with passenger demand by link;

Use a travel demand model to create a more accurate BRT demand estimate;

Estimate the mode shift to BRT;

Determine the utility of the BRT corridor;

Quantify the risk and uncertainty of the demand analysis.

2

3

4

5

6

UNDERSTAND THE EXISTING PUBLIC TRANSPORT NETWORK

Everett lacks a single “key” bus route, which the MBTA defines as a route that has high ridership and higher frequency standards (see Appendix A). It is currently served by nine routes with confusing, infrequent service that prioritize coverage over frequency. None of the existing bus routes provides a fast connection to the job centers of downtown Boston, Cambridge, or Somerville. These routes are not well-coordinated, and northbound passengers have to choose a transfer point, all of which have less service.

Existing bus routes on Broadway include the 104, 105, and 109, which all terminate at Sullivan Station. MBTA’s Better Bus Project has identified routes 104 and 109 as important routes in the system that are frequently crowded and need more service.8 There are additional routes that serve parts of Broadway and then run to Wellington station, providing a similar connection to Boston.

It is impossible to get to most of the region from Everett by transit without making a transfer, since the core of regional jobs in Boston and Cambridge can only be reached by changing to the Orange Line or at Sullivan for a bus to Cambridge. More than half of passengers boarding a bus in Everett are heading to downtown Boston (or beyond) and have to transfer. From Sullivan Square, passengers can also transfer to the 92 and 93 and continue to Haymarket via local service in Charlestown or to several routes to access Somerville and Cambridge.

DAILY PASSENGERS PER BUS ROUTE

The MBTA bus routes operating along Broadway in Everett collectively carry more than 1,500 weekday passengers per direction, including to both Sullivan and Wellington stations.9 North Washington Street from Charlestown, across the bridge to Haymarket Square, also carries at least 1,500 bus passengers per direction each weekday. More than 10,000 passengers pass through Sweetser Circle every day, with routes fanning out north on Main Street and Broadway, south on Broadway, and west on Revere Beach Parkway.

BUS FREQUENCY PER ROUTE

There is a bus on the upper portion of Broadway in Everett on average every four or five minutes toward Boston at peak times. At peak hour, including buses operating on Main Street, there are approximately 25 buses per hour from Everett to Sullivan and Wellington spread across these eight routes. However, these buses are split over a number of routes with multiple destinations and transfer points, and the schedules are not well coordinated, so longer waits are common.

BUS AND OTHER VEHICLE SPEEDS

Everett’s existing bus service is plagued by congestion and slow speeds. Today the 104/105/109 bus has an average operating speed of 7 mph during peak hours and a scheduled speed of 13 miles per hour at off-peak times.

Vehicle speeds along the corridor depend on the portion of the corridor as well as congestion. Between Glendale Square and Sweetser Circle, vehicle speeds range from 15 mph to 20 mph at off-peak times down to 8 mph to 12 mph at peak time, not including congestion getting through Sweetser Circle, where just traversing the circle can take several minutes. Between Sweetser Circle and Sullivan Square, traffic speeds can range from 30 mph at off-peak times to less than 10 mph when the area is congested.

BOARDINGS AND ALIGHTINGS BY STOP

MBTA boarding and alighting data by bus stop of each bus route operating along the Everett‒Boston corridor can be combined to calculate the number of onboard passengers along each road segment. These can be aggregated for all of the routes in the corridor to determine the maximum passenger load on the critical link.

EXISTING ORIGINS AND DESTINATIONS OF PASSENGER VEHICLE TRIPS

In order to inform our recommendations about BRT service and infrastructure between Everett and Boston, ITDP directed a study undertaken by AECOM to identify travel patterns between the two cities and specifically travel patterns along Rutherford Avenue. This included weeklong traffic counts at three separate points along the corridor, as well as zone data from Streetlight, which tracks cell phone data along the corridor. The main conclusions are summarized below and a detailed description of the analysis is provided in appendices D and E.

Approximately 95% of daily trips from Everett are made in passenger vehicles, but these trips have destinations across the region: Only a relatively small portion are to destinations in Boston and Cambridge that would be served by BRT. Of vehicles starting in Everett and crossing the Alford Street Bridge to Charlestown (10,370), slightly more are headed to Cambridge/Somerville (4,600) than to downtown Boston (3,600). Expanding bus service to provide convenient and efficient

connections between Everett and Cambridge/Somerville, Charlestown, and downtown Boston could encourage more transit use to these areas. In addition, providing transit from Everett to other transit connections offers more flexibility to travel to areas that were not previously accessible.

EXISTING ORIGINS, DESTINATIONS, AND TRANSFERS OF TRANSIT TRIPS

It is important to investigate the destinations and transfers of Everett transit passengers, since where they are going from Everett will impact the efficacy of the BRT, especially if routes are extended beyond Sullivan. To do this, ITDP used data from the MBTA’s Origin-Destination-Transfer (ODX) data set, developed by MIT.10 We used a typical time of year and looked at the relative prevalence of destinations and transfers for transit passengers with origins in Everett and their destinations (see Appendix F). Based on these data, the following trends are observed.

Approximately one in four passengers boarding a bus in Everett has a destination in downtown Boston for which they currently have to make a transfer at Sullivan but where a direct bus would create a one-seat ride.

However, the transfer at Sullivan is important, because there are an

additional one in three passengers going beyond downtown Boston on the Orange Line (about 20% of all riders) or to Cambridge and Somerville

(about 12%) who would likely make use of the transfer points there. Ridership in Cambridge and Somerville is quite distributed; there is no single destination point or bus route that has significantly more ridership than any other, so it is difficult to recommend a specific destination to serve with a new BRT route from Everett. Such a route would be used by additional travelers transferring from the Orange Line, however, which may suggest a specific desire line. If a route to Cambridge were selected, it is likely that ridership along this route would increase.

ESTIMATE THE UTILITY OF THE BRT CORRIDOR

It is important to examine whether the proposed BRT will save passengers time and money compared to the existing public transport system. Comparing the travel time and costs for a given set of trips before and after the BRT will shed some light on this and give an indication of how much utility the new BRT corridor would provide potential passengers. By converting travel time to a monetary value, we can examine how much changes to the transport system might encourage people to switch modes. ITDP compared travel time and costs for trips between Everett Square and either Haymarket or Back Bay, with and without the BRT.

UTILITY OF DIRECT EVERETT BRT SERVICE TO HAYMARKET FOR PASSENGERS DESTINED FOR HAYMARKET

While the total travel times are the same, currently Everett transit passengers heading to Haymarket have to transfer to the Orange Line at Sullivan Square, while drivers do not (Table 1). For those with limited mobility—whether that’s people using wheelchairs, pushing a stroller, or carrying luggage—a transfer can be a significant physical burden in addition to taking extra time. The driving trip is three times more expensive than transit because of the cost of parking in downtown Boston (unless that

cost is subsidized by an employer). Direct BRT service between Everett and Haymarket eliminates the transfer and parking cost and is estimated to take the same amount of time as driving. If it were possible to reduce the BRT

trip from Everett to Haymarket to less than 35 minutes, it would be even more competitive with driving.

Table 1: Comparing Cost of Peak-Hour Trip from Everett to Haymarket on Different Modes

In Vehicle Time Transfer + Walk + Walk Time Total Time Cost of Time11 (A) Trip Cost12 (B) Parking Cost (C) Total Cost (A+B+C)

Existing Bus + Train 24 min13 11 min 35 min $10.50 $4.80 $0 $15.30

Car

BRT 35 min14 0 min15 35 min16 $10.50 $5.75 $3017 $46.25

29 min 6 min 35 min $10.50 $3.40 $0 $13.90

UTILITY OF DIRECT EVERETT BRT SERVICE TO HAYMARKET FOR PASSENGERS DESTINED FOR DOWNTOWN CORE

For this analysis, ITDP assumed that continuing the Everett‒Boston BRT route past Haymarket, through the streets of downtown Boston, with speeds that are competitive with the (grade-separated) Orange Line would be difficult. However, transit riders perceive in-vehicle travel time differently from waiting time (which is how transfer time is generally measured), so there are some marginal benefits to Everett passengers of a one-seat transit ride to downtown Boston. That route extension could be explored further, but here it is assumed that the BRT route would terminate at Haymarket.

Direct BRT service between Everett and Haymarket would not eliminate the transfer for passengers destined for other areas farther into the core of downtown, such as Back Bay. Passengers would still need to transfer from the BRT to the Orange Line. Making this transfer farther upstream from Haymarket—at Wellington or Sullivan—makes for a faster trip, since even with the most aggressive transit priority on a BRT corridor, the grade-separated Orange Line will be faster than a parallel surface route (see Table 2).

11 This estimate uses a time value of $18/hour, (VTPI 2020a) and adjusts it for income in Massachusetts as compared to the national average, plus inflation since the study, to $17.94, which was rounded to $18 for simplicity. For transit, the current transit fares are assumed. For driving hyperlink 2020 IRS reimbursement rates are used General Transit Feed Specification (GTFS) estimated travel time. Estimated, based on pre-COVID traffic. This assumes no walk time from the parking garage to the car driver’s destination, which is fair, given ample parking adjacent to Haymarket. Estimated, based on pre-COVID traffic. Parking cost was estimated from a 2019 survey of Boston parking garages (Kennedy 2019).

Table 2: Comparing Cost of Peak Hour Trip from Everett to Back Bay on Different Modes

In Vehicle Time Transfer + Walk + Walk Time Total Time Cost of Time18 (A) Trip Cost19 (B) Parking Cost (C) Total Cost (A+B+C)

Existing Bus + Train (transfer at Sullivan)

BRT + train (transfer at Haymarket)

BRT + train (transfer at Sullivan)

Car 33 min 11 min 44 min $13.20 $4.80 $0 $18.00

38 min 11 min 49 min $14.70 $4.80 $0 $19.50

31 min 11 min 47 min $14.10 $4.80 $0 $18.90

45min20 -- 45min21 $13.50 $6.90 $3022 50.40

This analysis of the travel time and costs of different modes highlights that people driving are already paying nearly triple the cost of taking transit. In planning an Everett‒Boston BRT corridor it is important to understand drivers’ willingness to pay for the utility of driving. The Central Transportation Planning Staff (CTPS) should utilize a full mode choice model to develop a more complete assessment of the utility of a new BRT service and a more refined estimate of the feasibility of the BRT system.

ROUGHLY ESTIMATE MODE SHIFT TO BRT

Transit use on bus routes in Everett declined less during the COVID-19 outbreak than overall transit use, indicating that transit is a lifeline for many Everett passengers. High-quality BRT service between Everett and Boston that provides real transit priority and speeds up the trip will elevate the bus experience for the thousands of people who currently use transit in Everett.

Shifting trips from motor vehicles to BRT to reduce air pollution and congestion may also be a goal of the BRT corridor project. Achieving this shift will depend on several factors including the relative speeds, throughput, and cost of driving and BRT (see tables 1 and 2). Everett’s low public transit mode share for all trip purposes (3% to 4%)23 suggests some potential to attract new transit riders with convenient and efficient new BRT service. Mode shift to BRT would depend on the cross-price elasticity of the

trips being taken. A reasonable estimate24 yields a 4% mode shift from cars to BRT, or 125 people per day between Everett and downtown Boston. New BRT service to Kendall Square would probably attract additional trips from cars.

ROUGHLY ESTIMATE BRT DEMAND

Demand on the future BRT corridor won’t exactly match the existing public transit demand for several reasons. Some bus routes or portions of routes may be eliminated with the introduction of the BRT service (see Service Planning section), and so these existing transit passengers may shift to alternative modes. If speeds on the BRT routes exceed existing bus speeds, BRT may attract passengers from other routes or modes. Likewise, extending BRT routes to new destinations not currently served by transit will induce demand.

The current transit demand on the corridor is 3,000 weekday passengers. Contributions to BRT demand can be roughly estimated as follows:

If 3% of that ridership is lost to other modes by eliminating portions of existing routes (-90 passengers/day);

Mode shift from private vehicles (+125 passengers/day) (see previous section);

And new BRT routes to Kendall attract 5% of the current driving trips from Everett to Cambridge (+150 passengers/day)

While these rough estimates need more interrogation, this back-of-theenvelope calculation suggests a daily demand of 3,185 weekday passengers on the BRT corridor. This does not consider land-use changes such as densification around BRT stations, or travel demand management measures such as reducing parking availability or increasing parking costs. The influence of those factors can be better determined in a full transport model.

REFINE BRT DEMAND AND MODE SHIFT ESTIMATES WITH FULL TRANSPORT MODEL

An integrated land-use and transport model can more accurately estimate future BRT demand in the corridor by taking into consideration people’s travel choices and the effect of congestion and land-use changes on different parts of the transport network. ITDP plans to work with the regional MPO on a regional transport modeling study that can produce a more accurate BRT demand analysis.

A logistic regression mode choice model can estimate how many Everett travelers will opt for the BRT, rail, private vehicle, bike, or walking. As part of this forthcoming modeling exercise, mode shift with a new Everett‒Boston BRT corridor will be examined.